This cross-sectional study used a semi-automated analysis technique to quantify regional brain cerebrospinal fluid (CSF) volumes derived from computed tomography (CT) in 84 healthy men ranging from 21 to 82 years of age and 28 patients meeting Research Diagnostic Criteria for alcohol dependence. The goals were to replicate an earlier CT study of an independent sample of alcoholic and control subjects (Pfefferbaum et al., 1988a; Zipursky et al., 1988) and to compare CT assessments of brain changes with magnetic resonance imaging (MRI) assessments made in the same alcoholic patients (Pfefferbaum et al., 1992). Regional brain changes associated with normal aging were derived by regression analysis, using CT data collected from the healthy control subjects. As in the earlier CT study and in the concurrent MRI study, ventricular and sulcal CSF volumes in alcoholic patients were greater than would be expected for their age. Furthermore, the present CT study replicated the previous CT and MRI findings of a positive relationship between age and CSF volume enlargement in alcoholic patients over and above the normal age-related increase in CSF volume, suggesting greater vulnerability of the aging brain to alcohol. Comparison of CT- and MRI-derived estimates of ventricular and cortical sulcal volume revealed high correlations (> 0.80). MRI and CT produced similar absolute ventricular volumes, while MRI produced larger sulcal volume estimates than did CT. The difference in sulcal volume estimate may be due to differences between CT and MRI in slice thickness and sensitivity to partial volume effects.